Shouldn't that be silicon rather than silicone?

Downvoted on a technicality. I think that as we reach the limits of what silicon can do for us, Moore's law will continue via some other kind of technology.

Upvoted because 20 years seems way too long. Interestingly, Kaku predicted the end of Moore's law for silicon "in 20 years" back in 2003. 2023 seems more likely to me than 2032.

However this is only true if you are applying "Moore's Law" in the stricter sense, which by definition only refers to silicon. These days "Moore's Law" is usually used to refer more broadly to increases in cost-performance for commercially available computers. My guess is that shortly after Moore's law for silicon starts flatlining, something else (probably molecular computers built around graphene will become available. It is far from obvious that the new substrate will improve at the same steady pace and overall speed as Moore's Law for silicone. It could go much faster or much slower, possibly in fits and starts as opposed to steady, gradual process as we've seen for decades.

Downvoted because of this "I believe in that Moore's law modeling silicone improvement will collapse in 20 years" but I would have upvoted instead if you had said that the collapse would be in the direction of flatlining within 20 years.